We Keep Coding

Implement background process in Spring

I need to implement Spring process which checks database table for new rows and makes calculations. I'm thinking to implement infinite loop which is triggered every 10 minutes.
Is there someway to implement this with Spring Boot? I can always use Java Thread but it's for sure it's better to let Spring to manage this.

You can try scheduling with Spring Schedule
Here is a official example
Technically, you enable scheduling with #EnableScheduling and annotated your task with #Scheduled(fixedRate=600000).
Another config you can used to tune your scheduler:
fixedRate: Execute the annotated method with a fixed period in milliseconds between invocations.
fixedDelay: Execute the annotated method with a fixed period in milliseconds between the end of the last invocation and the start of the next.
cron: A cron-like expression, extending the usual UN*X definition to include triggers on the second as well as minute, hour, day of month, month and day of week.

find the below sample code
import org.springframework.scheduling.annotation.Scheduled;
import org.springframework.stereotype.Component;
#Component
public class AppScheduler {
#Scheduled(fixedRate = 10000)
public void myScheduler() {
System.out.println("Test print");
}
}

It looks like this questions is old, but I would like to answer.
See, you can make an object of ThreadPoolTaskExecutor and assign the background process to it.
Here is my detailed code if anyone wants to see.
https://github.com/xbox2204/SpringBoot-JPA-Swagger-Angular
First thing to do would be, add these two annotation in your application starter class.
#EnableAsync
#EnableScheduling
Now, create you TaskExceutor in the same class and assign it to bean with a name, so that background tasks can be created anywhere in your application and attached with this bean.
#Bean(name="thPlExectr")
public Executor getExecutor(){
return new ThreadPoolTaskExecutor();
}
Now create a component class somewhere in the project, where you you will create the background task.
Here, you will mention the frequency with which you want your task to be exceuted. I wanted to print a statement every 5 second, you can choose your own frequency and give your own method definiton. Also, make it async and attach it with the TaskExecutor bean mentioned above, so that it doesn't interrupt the normal flow of your application.
#Component
public class BackgroundTasks {
private static final Logger log= LoggerFactory.getLogger(BackgroundTasks.class);
#Async("thPlExectr")
#Scheduled(fixedRate = 5000)
public void backTask(){
log.info("Hi Vineet! I am a background task");
}
}

Spring - Dynamically add and remove scheduled task

I am working on spring based server application. Basically it will poll scores of various sporting events in very short interval and save in db. For polling there will be many(can be around 100) calls to different apis concurrently at regular interval for example some api call will have 3 seconds interval some have 5 seconds etc., server will keep polling for latest data at frequent interval.
These calls will be added and removed dynamically. I have little experience in using spring. I think I have to use some scheduler. Can anyone point in right direction what approach or which scheduler is best in this scenario.

In essence you want to inject an instance of a scheduling task executor
#Configuration
public class MyApplicationConfiguration {
#Bean
public ThreadPoolTaskScheduler threadPoolTaskScheduler() {
ThreadPoolTaskScheduler tpts = new ThreadPoolTaskScheduler();
// maybe configure it a little?
return tpts;
}
#Bean
public MyService myService() {
return new MyService();
}
}
class MyService {
#Autowired
private ThreadPoolTaskScheduler tpts;
public void doSomething() {
Runnable task = ...
tpts.scheduleWithFixedDelay(task, 1000);
}
}
You can see a reasonable guide here, or the SchedulingTaskExecutor Javadoc and the Spring Task Execution and Scheduling Reference

You can use #Scheduled Spring Annotation for this. Refer this link for examples.

Spring schedule cron does not work with hours [duplicate]

How can I configure the time zone for a Spring based #Scheduled cron job?
Background:
I have a job that executes once a day, say 2 PM, using Spring's #Scheduled annotation:
#Scheduled(cron = "0 0 14 * * *")
public void execute() {
// do scheduled job
}
The problem is that 2 PM differs between different servers, because Spring uses on TimeZone.getDefault() internally. Moreover, the JavaDoc of TimeZone.getDefault() states that:
Gets the default TimeZone for this host. The source of the default TimeZone may vary with implementation.
In other words, the time zone is not determined. It may depend on JVM implementation, server time zone configuration, server location, and / or other unknown factors. Consequently, the cron job triggers on different times on different servers, unless there is a way to explicitly set which time zone that should be used?
I am using Spring 3.2.2.
Update
As of Spring 4, Spring Jira issue SPR-10456 has been resolved. Consequently, the #Scheduled annotation has a new zone attribute for exactly this purpose.

It turned out that I could not use the #Scheduled annotation, but I implemented a work-around. In the JavaDoc of the SchedulingConfigurer it is stated that:
[SchedulingConfigurer is] Typically used for setting a specific TaskScheduler bean to be used when executing scheduled tasks or for registering scheduled tasks in a programmatic fashion as opposed to the declarative approach of using the #Scheduled annotation.
Next, I changed the cron job to implement the Runnable interface and then updated my configuration file to implement the SchedulingConfigurer, see below:
#Configuration
#EnableScheduling
#ComponentScan("package.that.contains.the.runnable.job.bean")
public class JobConfiguration implements SchedulingConfigurer {
private static final String cronExpression = "0 0 14 * * *";
private static final String timeZone = "CET";
#Autowired
private Runnable cronJob;
#Bean
CronTrigger cronTrigger() {
return new CronTrigger(cronExpression, TimeZone.getTimeZone(timeZone));
}
#Override
public void configureTasks(ScheduledTaskRegistrar taskRegistrar) {
taskRegistrar.addCronTask(new CronTask(job, cronTrigger()));
}
}
Please read the JavaDoc of the #EnableScheduling for more information.
Update
As of Spring 4, Spring Jira issue SPR-10456 has been resolved. Consequently, the #Scheduled annotation has a new zone attribute for exactly this purpose, e.g.
#Scheduled(cron = "0 0 14 * * *", zone = "CET")
public void execute() {
// do scheduled job
}

There is element zone in annotation #Scheduled, starting from version 4.0.
You can insert a timezone as a string that can be accepted by java.util.TimeZone.

Your code should be like this:
#Scheduled(cron = "0 0 14 * * *", zone = "GMT-5")
public void execute() {
// do scheduled job
}
"Zone" is gonna be the desired country's timezone.
Here is a nice tutorial about scheduled tasks with Spring:
https://www.baeldung.com/cron-expressions

You can also use time zone with #Scheduled tag in spring-boot like this :
#Scheduled(cron = "0 0 14 * * *" , zone = "GMT+5:00")
public void execute() {
// do the scheduled job
}

I doubt you want different jobs or parts of application to use different time zones. Assuming you want to have it all consistent and DRY, either configure OS on all servers to have consistent time zone, or set user.timezone Java system property for all of the application servers. Centrally manage configuration (OS, application server), and for that puppet and chef can be very useful.

how to suspend job in quartz scheduler?

Hi i am creating an application that executes the method of a class based on a cron expression. For that i am using spring quartz in which i have to configure all these stuffs in my spring-file it works fine and jobs are executing based on the cron expression, But now i want to suspend the next execution of particular job in java class based on the choice of a user from UI. Then is there any way to do this ??
can i get the details of all running job it the context ? if so then i can filter the jobs and try to suspend that job for next execution.

Inject your SchedulerFactoryBean. Use it's getScheduler method to obtain a quartz Scheduler and use rescheduleJob or other methods from quartz API to perform your task.

I got it work by use of following code
stdScheduler is a scheduleFactoryBean
String groupnames[] = stdScheduler.getJobGroupNames();
for (String groupName : groupnames) {
String[] jobNames = stdScheduler.getJobNames(groupName);
for (String jobName : jobNames) {
if (jobName.equals("jobName")) {
stdScheduler.pauseJob(jobName, groupName);
}
}
}

We can use the stdScheduler.pauseJob(JobKey jobKey) method to reduce the number of loops mentioned in the code above

If you got hold of the SchedulerFactoryBean by injection or somehow else there are the convenience methods:
schedulerFactoryBean.getScheduler().standby();
schedulerFactoryBean.getScheduler().start();
When using Quartz directly also this works:
#Autowired
private Scheduler scheduler;
...
scheduler.standby();
...
scheduler.start();

Running a Job only once Using Quartz

Is there a way I could run a job only once using Quartz in Java? I understand it does not make sense to use Quartz in this case. But, the thing is, I have multiple jobs and they are run multiple times. So, I am using Quartz.
Is this even possible?

You should use SimpleTrigger that fires at specific time and without repeating. TriggerUtils has many handy methods for creating these kind of things.

Yes, it's possible!
JobKey jobKey = new JobKey("testJob");
JobDetail job = newJob(TestJob.class)
.withIdentity(jobKey)
.storeDurably()
.build();
scheduler.addJob(job, true);
scheduler.triggerJob(jobKey); //trigger a job inmediately

In quartz > 2.0, you can get the scheduler to unschedule any job after work is done:
#Override
protected void execute(JobExecutionContext context)
throws JobExecutionException {
...
// process execution
...
context.getScheduler().unscheduleJob(triggerKey);
...
}
where triggerKey is the ID of the job to run only once. After this, the job wouldn't be called anymore.

Here is an example of how to run a TestJob class immediately with Quartz 2.x:
public JobKey runJob(String jobName)
{
// if you don't call startAt() then the current time (immediately) is assumed.
Trigger runOnceTrigger = TriggerBuilder.newTrigger().build();
JobKey jobKey = new JobKey(jobName);
JobDetail job = JobBuilder.newJob(TestJob.class).withIdentity(jobKey).build();
scheduler.scheduleJob(job, runOnceTrigger);
return jobKey;
}
see also Quartz Enterprise Job Scheduler Tutorials → SimpleTriggers

I'm not sure how much similar is Quartz in Mono and Java but this seems working in .Net
TriggerBuilder.Create ()
.StartNow ()
.Build ();

I had to ask myself if it made sense to try to configure a job and add checks if it had been run already as suggested in Marko Lahma's answer (since scheduling a job to run once results in it being run once, every time we start the app). I found examples of CommandLineRunner apps which didn't quite work for me, mostly because we already had an ApplicationRunner which was used for other jobs which use Quartz scheduling / cron. I wasn't happy with having Quartz initialize this job using a SimpleTrigger, so I had to find something else.
Using some ideas from the following articles:
Multiple Spring boot CommandLineRunner based on command line argument
Run Spring Batch Job programmatically?
Firing Quartz jobs manually
Is there any way to get job keys in Quartz by job name
How to list all Jobs in the Quartz Scheduler
Spring Boot CommandLineRunner and ApplicationRunner
I was able to piece together a working implementation which allows me to do the following:
run existing jobs via Quartz, on a timer
run new job, one time programmatically (single use Quartz job using the SimpleTrigger didn't satisfy my requirements, since it would be run once on every application load)
I came up with the following CommandLineRunner class:
public class BatchCommandLineRunner implements CommandLineRunner {
#Autowired
private Scheduler scheduler;
private static final Logger LOGGER = LoggerFactory.getLogger(BatchCommandLineRunner.class);
public void run(final String... args) throws SchedulerException {
LOGGER.info("BatchCommandLineRunner: running with args -> " + Arrays.toString(args));
for (final String jobName : args) {
final JobKey jobKey = findJobKey(jobName);
if (jobKey != null) {
LOGGER.info("Triggering job for: " + jobName);
scheduler.triggerJob(jobKey);
} else {
LOGGER.info("No job found for jobName: " + jobName);
}
}
}
private JobKey findJobKey(final String jobNameToFind) throws SchedulerException {
for (final JobKey jobKey : scheduler.getJobKeys(GroupMatcher.jobGroupEquals("DEFAULT"))) {
final String jobName = jobKey.getName();
if (jobName.equals(jobNameToFind)) {
return jobKey;
}
}
return null;
}
}
In one of my configuration classes I added a CommandLineRunner bean which calls the custom CommandLineRunner I created:
#Configuration
public class BatchConfiguration {
private static final Logger LOGGER = LoggerFactory.getLogger(BatchConfiguration.class);
#Bean
public BatchCommandLineRunner batchCommandLineRunner() {
return new BatchCommandLineRunner();
}
#Bean
public CommandLineRunner runCommandLineArgs(final ApplicationArguments applicationArguments) throws Exception {
final List<String> jobNames = applicationArguments.getOptionValues("jobName");
LOGGER.info("runCommandLineArgs: running the following jobs -> " + ArrayUtils.toString(jobNames));
batchCommandLineRunner().run(jobNames.toArray(ArrayUtils.EMPTY_STRING_ARRAY));
return null;
}
}
Later, I am able to initiate these jobs via the CLI without affecting my current Quartz scheduled jobs, and as long as no one runs the command via CLI multiple times, it will never be run again. I have to do some juggling of types since I accept ApplicationArguments, and then convert them into String[].
Finally, I am able to call it like this:
java -jar <your_application>.jar --jobName=<QuartzRegisteredJobDetailFactoryBean>
The result is that the job is initialized only when I call it, and it is excluded from my CronTriggerFactoryBean triggers which I used for my other jobs.
There are several assumptions being made here, so I'll try to summarize:
the job must be registered as a JobDetailFactoryBean (e.g.: scheduler.setJobDetails(...))
everything is essentially the same as a job with CronTriggerFactoryBean, excepting the lacking scheduler.setTriggers(...) call
Spring knows to execute the CommandLineRunner classes after the application has booted
I hardcoded the parameter being passed into the application to "jobName"
I assumed a group name of "DEFAULT" for all jobs; if you want to use differing groups this would need to be adjusted when fetching JobKey, which is used to actually run the job
there is nothing which prevents this job from being run multiple times via CLI, but it was triggered on every application load using SimpleTrigger approach, so this is better for me; if this is not acceptable, perhaps using StepListener and ExitStatus, etc. can prevent it from being executed twice

Another solution: There is a method .withRepeatCount(0) in SimpleSchedulerBuilder:
public final int TEN_SECONDS = 10;
Trigger trigger = newTrigger()
.withIdentity("myJob", "myJobGroup")
.startAt(new Date(System.currentMillis()+TEN_SECONDS*1000)
.withSchedule(SimpleScheduleBuilder.simpleSchedule()
.withRepeatCount(0)
.withIntervalInMinutes(1))
.build();